My thoughts on tires for GT6

[LeadFootLiam]

I agree with this. It would make tuning cars much more interesting, and more variety of setups that people can have.

 

[shawndre]

I wish that tires didn't "reset" after each race. It would be more strategic to decide whether to race on some used tires or buy another set. For example sports hards would last 10 races compared to racing softs lasting 20 laps.

 

[another_jakhole]

I wish that tires didn't "reset" after each race. It would be more strategic to decide whether to race on some used tires or buy another set. For example sports hards would last 10 races compared to racing softs lasting 20 laps.

It has potential, I guess. I think it might be more fit to be a feature in an open-world map. to a house, garage, puppy named satan that tears your leather seats apart... 👍


Seriously, it would be a great (or maybe obvious) feature for free roam.

 

[LVracerGT]

I don't doubt he doesn't know his stuff, but in don't agree with his opinion that the tire system is ok as it is. Don't get me wrong, the physics in GT5 are definitely the best to date in the series, as well as the best on a console. But they definitely need to be updated/redone. They want to be the real driving simulator, then use more realistic physics supported by real world data. I.E. iRacing or GTR2

 

[doyley101]

I don't doubt he doesn't know his stuff, but in don't agree with his opinion that the tire system is ok as it is. Don't get me wrong, the physics in GT5 are definitely the best to date in the series, as well as the best on a console. But they definitely need to be updated/redone. They want to me the real driving simulator, then use more realistic physics supported by real world data. I.E. iRacing or GTR2

I agree, it's overestimated how much more difficult realistic physics are. For instance, if it wasn't for the stages, I feel that RBR would be a lot easier than Colin Mcrae rally because the car reacts how you expect it to.

I don't think it would put younger people of either, my favourite games aged 10/11 were RBR, GTR2 and TOCA!

 

[Griffith500]

There was a lot of nonsense to sift through here, but here's my response.

While I agree that tire manufacturers aren't necessary as long as the tire model is advanced, it can only be good for the series. As far as tire width being far too much data to collect, I call BS.

PD is a Sony funded studio with NOTHING but time and money. If other console racers far more barbaric (in terms of technology when it was released and on what system, as Ridox stated above) there is NO excuse for PD to not have the time or resources.

Anyone who thinks using a simple grip multiplier is better than using real world ture data is 🤬 stupid and needs to go play NFS. There is far more to tires and how they perform than simple grip multipliers. Go play iRacing and tell me a top of the line world class tire model with manufacturer real world tire data doesn't make a difference.

GT needs a better tire model. Period. Anyone who thinks otherwise doesn't care about the advancement of this series. And given the leaked specs of the new consoles coming out (being on par with high end PCs) there is no reason GT shouldn't feature a true to life tire model with tread width and tire pressures being taken into account (and hopefully adjustable) because just these two values alone can make the tires 10 times more realistic than what we currently have. Tire manufacturers with options would be an added bonus.

Don't you realise that all tyre data are "grip multipliers" (that's all the physics engine needs from one tick to the next), or did you sift that out as nonsense? Maybe you ought to concentrate a bit more and try to see what I actually wrote, rather than what you expected I would.

I think it's hilarious that you mention iRacing's tyre model, which is, in terms of direction, exactly the kind of model I was talking about! Nowhere did I state that GT's tyre model is OK as it is, only that "collecting data" isn't enough - iRacing's model is largely physically-based, validated by empirical (experimentally measured, real-world) data collected in experiments designed by Dave Kaemmer himself, not from what is typically made available by tyre manufacturers - those multipliers aren't used directly anywhere in the sim, it calculates its own on the fly (to a degree, based on lots of pre-calculations, I'd guess). Good reading on the subject here, and an "interview" with Kaemmer here.

 

[LVracerGT]

Things like changing tread width are really only usable with a detailed physical (i.e. non-empirical) model, there's just far too much data to collect, unless you're going to blur between a few key real-world tyres, which isn't guaranteed to be any more accurate. And anyone who thinks that using real tyre data (which PD probably already do) is somehow automatically better than a simple "grip multiplier" is deluded - that's how all tyre models work, because that's fundamentally how simple friction models work - it's all "grip multipliers".

Here, you state that using changing tread width data wouldn't be more accurate and that anyone that thinks using real tire data is better than a simple grip multiplier is deluded. I'm sorry Griffith, I realize you may be a respected member on here, but this is 100% wrong.

First, changing tread width is a huge aspect. It can greatly change the way a car handles. Why do you think GM put wider tires on the back of the Camaro ZL-1? To put more grip in the rear.

Second, tire models are not just simple grip multipliers. I can see where you get this from, as softer or harder compounds can be adjusted with simple grip mulitpliers. But you're not taking into account what real world tire data can give. Such as slip angle coefficient, and sidewall spring rate/flex as a couple examples. Both can greatly affect how the tire performs and is different from tire to tire.

 

[Griffith500]

Here, you state that using changing tread width data wouldn't be more accurate and that anyone that thinks using real tire data is better than a simple grip multiplier is deluded. I'm sorry Griffith, I realize you may be a respected member on here, but this is 100% wrong.

First, changing tread width is a huge aspect. It can greatly change the way a car handles. Why do you think GM put wider tires on the back of the Camaro ZL-1? To put more grip in the rear.

Second, tire models are not just simple grip multipliers. I can see where you get this from, as softer or harder compounds can be adjusted with simple grip mulitpliers. But you're not taking into account what real world tire data can give. Such as slip angle coefficient, and sidewall spring rate/flex as a couple examples. Both can greatly affect how the tire performs and is different from tire to tire.

You forgot the part where you actually read what I wrote, and threw away whatever prejudices you may be harbouring, for whatever reason. I never said that a tyre model shouldn't allow different tread widths, I said that a tyre model should be able to represent any tyre width, height, compound etc. regardless of whether it exists and can be "measured". I'm simply arguing in favour of physical models rather than empirical ones.

It's strange that you use the word "coefficient" and state that it's somehow superior to a mere "multiplier", seemingly ignorant that they're synonyms. I know it's hard given the culture we all come from, but please do try to separate the emotional component of a word's connotation from its actual meaning.
The force generated by a tyre is primarily due to its time-varying "coefficient of friction" as described in the Wikipedia link I gave in the first instance. Coefficient of friction = grip multiplier - therefore all tyre models require grip multipliers, hence my chiding of anyone who seems to think "grip multipliers" are somehow inherently bad. I'm simply talking about methods to acquire these multipliers according to what the tyre is doing in real time, rather than worry about trying to interpolate within an incomplete and usually unsuitable dataset. Don't forget that "slip angle" is a made-up property that was only conceived to make analysing results of tyre testing easier, and its cousin "slip ratio" is the reason GT5 needs a low speed hack in the first place.

I suggest you actually listen to what Dave Kaemmer has to say on the subject, because I couldn't care less what my "reputation" is (which I'm sure isn't that great anyway, and isn't a reason to believe what I say without checking the facts behind what's being discussed), but his surely isn't in question here.
Add to that the fact that T10 needed to get Pirelli to do custom tests just so they could tidy up their own model (which was actually only problematic because they were using multiple manufacturers' data), and you see the scale of the challenge of the empirical approach here.

Obviously, real data is needed to validate any physical model that may be developed, but these tests will be very different from the sorts tyre companies typically make, and need only be done to cover the broad range of tyres available, based on the model's "inputs". The empirical methods require that every single one be tested for absolute accuracy. Obviously that's not possible, so you'll be measuring specific tyres in a range anyway, with no real guarantee that anything that falls in between those tyres in the range is at all accurate, which is what I was saying above. Maybe I mixed up my terminology slightly.
Once the physical model is "complete", it may in the future be a simple case of specifying the tyres' dimensions and materials, and the model does the rest, without having to bother the tyre companies for data from destructive tests too often.

 

[LVracerGT]

You forgot the part where you actually read what I wrote, and threw away whatever prejudices you may be harbouring, for whatever reason. I never said that a tyre model shouldn't allow different tread widths, I said that a tyre model should be able to represent any tyre width, height, compound etc. regardless of whether it exists and can be "measured". I'm simply arguing in favour of physical models rather than empirical ones.

It's strange that you use the word "coefficient" and state that it's somehow superior to a mere "multiplier", seemingly ignorant that they're synonyms. I know it's hard given the culture we all come from, but please do try to separate the emotional component of a word's connotation from its actual meaning.
The force generated by a tyre is primarily due to its time-varying "coefficient of friction" as described in the Wikipedia link I gave in the first instance. Coefficient of friction = grip multiplier - therefore all tyre models require grip multipliers, hence my chiding of anyone who seems to think "grip multipliers" are somehow inherently bad. I'm simply talking about methods to acquire these multipliers according to what the tyre is doing in real time, rather than worry about trying to interpolate within an incomplete and usually unsuitable dataset. Don't forget that "slip angle" is a made-up property that was only conceived to make analysing results of tyre testing easier, and its cousin "slip ratio" is the reason GT5 needs a low speed hack in the first place.

I suggest you actually listen to what Dave Kaemmer has to say on the subject, because I couldn't care less what my "reputation" is (which I'm sure isn't that great anyway, and isn't a reason to believe what I say without checking the facts behind what's being discussed), but his surely isn't in question here.
Add to that the fact that T10 needed to get Pirelli to do custom tests just so they could tidy up their own model (which was actually only problematic because they were using multiple manufacturers' data), and you see the scale of the challenge of the empirical approach here.

Obviously, real data is needed to validate any physical model that may be developed, but these tests will be very different from the sorts tyre companies typically make, and need only be done to cover the broad range of tyres available, based on the model's "inputs". The empirical methods require that every single one be tested for absolute accuracy. Obviously that's not possible, so you'll be measuring specific tyres in a range anyway, with no real guarantee that anything that falls in between those tyres in the range is at all accurate, which is what I was saying above. Maybe I mixed up my terminology slightly.
Once the physical model is "complete", it may in the future be a simple case of specifying the tyres' dimensions and materials, and the model does the rest, without having to bother the tyre companies for data from destructive tests too often.

I don't think we were on the same page, I agree with this post and I apologize if I offended you. It seemed like you were stating that grip multipliers were what made a tire model and that it was a good thing the way GT5 is using it.

From what I understand, GT uses a multiplier that increases equally with each time compound. Rough example if comfort soft = 1 sports hard would be 1.5, sports medium 2, etc.

I've listened to Dave's interview a while back but I'll have to have another listen.

 

[McOwnage]

They should also include drag slicks for the drag racers. And I disagree with the thought that manufactures should be included. Personally I think they should have category's like they do, just not calling them sports hard, etc. They should just have street tires, sport tires, racing tires, and drag slicks. Possibly with 3 choices of manufactures if they want to include them.

 

[Griffith500]

I don't think we were on the same page, I agree with this post and I apologize if I offended you. It seemed like you were stating that grip multipliers were what made a tire model and that it was a good thing the way GT5 is using it.

From what I understand, GT uses a multiplier that increases equally with each time compound. Rough example if comfort soft = 1 sports hard would be 1.5, sports medium 2, etc.

I've listened to Dave's interview a while back but I'll have to have another listen.

It might seem a bit simplistic, but the tyres are fictional. Having various compounds offer gradually more grip is actually nice for balance. What's problematic for GT5 is that every car seems to have the same steady-state grip on a given compound regardless of tyre construction and whatever else matters there (that may be within the realms of realism, but I've not looked at real world tests of different cars on the exact same tyres). I personally think that's deliberate, and it's been discussed that the finer tyre variables are actually handled per-car - the F1 cars might be a good indicator of this given the split of character usually evident in Comforts and Sports is reportedly absent there. That that character can usually be felt at all points to something more subtle going on anyway - i.e there are normally three main types of tarmac tyres with their own distinct driving feel, and each with different compound options.

Expanding that range is important, but it's potentially far easier and just as accurate overall, not mention enjoyable, just to use generic tyres for that at first. There's no real need to use specific models of branded tyres, because they're only applicable to a handful of cars anyway - on the other hand, if it's practical, there's no real reason not to at first, either. In the end, though, it really needs to be a standalone model that can be fed any parameters to mimic any tyre that has ever been made, especially given how most of those can probably never be tested properly. It'd be great to put cross-plys on the GT40, for instance (which may even be accounted for in the per-car tyre physics parameters for that car anyway, assuming you don't use racing tyres on it.)

 

[Stephanos82]

Nice thread, to bad I got here too late.. Why does everyone mention iRacing as an example for the best tyre physics? Having started racing in real life with my sweet mx5 1.8 1994 and moving up quite slowly I can't even touch iRacing anymore.. It just feels like driving on ice
Rfactor2 is becoming better by the update but IMO I don't think there's a SIM that can compare to LFS at the moment! It's difficluto to say how they did it, but they totally captured the feeling of weight transfer, how the tyres compress, flex and react .. I'm no physics techie, I just know it feels right!
Cars in GT5 have no sense of being inflated, no suspension movement worth talking about etc.. Forza 4 feels a lot better in comparison :/
Let's all hope that GT6 we'll fix all that! +1 to LVracerGT and Griffith

 

[TorqueHappenS08]

Hellow,And welcome to GTPlanet, Please try not to bump old threads.

https://www.gtplanet.net/forum/showthread.php?t=282389

 

[SuperCobraJet]

Cars in GT5 have no sense of being inflated, no suspension movement worth talking about etc.. Forza 4 feels a lot better in comparison :/

Many of us who have played GT5 and Forza 4 have expressed the same sentiments throughout the threads here.

Let's all hope that GT6 we'll fix all that! +1 to LVracerGT and Griffith

It sure sounds like Kaz is working on doing just that, from the collaborative effort he is using to formulate the GT6 physics model.

 

[Streeto]

I'm almost sure they're going to change the comfort to racing idea.

The new tire model and working with yokohama will of seen to that.

We might even get branded tires (yokohama most likely)...